Innovative Metrics for Reporting and Comparing the Glycan Structural Profile in Biotherapeutics

Abstract

Glycosylation is a critical quality attribute in biotherapeutics, impacting properties such as protein stability, solubility, clearance rate, efficacy, immunogenicity, and safety. Due to the heterogenic and complex nature of protein glycosylation, comprehensive characterization is demanding. Moreover, the lack of standardized metrics for evaluating and comparing glycosylation profiles hinders comparability studies and the establishment of manufacturing control strategies. To address both challenges, we propose a standardized approach based on novel metrics for a comprehensive glycosylation fingerprint which greatly facilitates the reporting and objective comparison of glycosylation profiles. The analytical workflow is based on a liquid chromatography-mass spectrometry-based multi-attribute method. Based on the analytical data, a matrix of glycosylation-related quality attributes, both at site-specific and whole molecule level, are computed, which provide metrics for a comprehensive product glycosylation fingerprint. Two case studies illustrate the applicability of the proposed indices as a standardized and versatile approach for reporting all dimensions of the glycosylation profile. The proposed approach further facilitates the assessments of risks associated with changes in the glycosylation profile that may affect efficacy, clearance, and immunogenicity.

Keywords: N-glycosylation; biosimilarity; biotherapeutics; characterization; glycan analysis; glyco-similarity; glycopeptide mapping; glycosylation index; standardization comparability.

Figure 1

Schematic workflow exemplifying the glycan analysis of a complex glycoprotein with three domains (A, B and C) containing two N-glycosylation sites in each domain. Domains B and C are partially glycosylated at the N1 site. Glycan analysis by (1) glycan release method by PNGase F: the site-specific information is lost; or by (2) glycopeptide mapping: the site-specific information is maintained.

Figure 2

An example of an IgG1 antibody fusion molecule containing 3 N-linked glycans in each heavy chain: one in the CH2 domain of the Fc region and two (N1 and N2) in the fused protein domain. Each site is characterized by its own glycosylation profile. Abbreviations used: VL and CL—variable and constant domains, respectively, of the light chain; VH and CH—variable and constant domains, respectively, of the heavy chain; CH1, CH2, and CH3—constant domains 1, 2, and 3, respectively, of the heavy chain.

Figure 3

Macro-heterogeneities resulting from the partial site occupancy of the N2 site (ordered according to their expected abundance.

Figure 4

Schematic structure of recombinant FSH.